Neurological injury after cardiac arrest – setting a case of prolonged re‐warming into a developing research context

Hypothermia is both a reversible precipitant of cardiac arrest and a well‐established neuroprotective mechanism. In 1975, Greipp et al. demonstrated induced hypothermia to be a safe, reproducible therapy to support complex cardiac and neurological surgery without cardiopulmonary bypass [1]. National Institute for Health and Care Excellence (NICE) guidance currently supports an active cooling strategy for comatose survivors of cardiac arrest, both in and out of the hospital, to maintain a core body temperature of 32–34 °C for 12–24 h after cardiac arrest to minimise brain injury [2]. However, if accidental hypothermia has precipitated a cardiac arrest, active rewarming will instead be required to restore circulation. In this issue of Anaesthesia Reports, Grewal and Thomas describe a case of prolonged cardiac arrest in a patient with accidental hypothermia who achieved a good neurological outcome with active re‐warming [3]. In this accompanying editorial, we consider the current evidence for temperature control interventions in the context of cardiac arrest.

Oxygen delivery and demand imbalance results in varying degrees of hypoxic tissue injury. A temperature reduction of 10 °C can reduce neurological oxygen demand by 50–75% [4]. Therefore, a subsequent reduction in blood flow (and oxygen delivery) may be similarly tolerated with minimal tissue damage.

Survival rates following cardiac arrest have remained largely unchanged in the last decade despite best efforts with a considerable proportion of patients succumbing to secondary brain injury as opposed to the original primary cardiac insult [5, 6]. As a result, interest in employing the protective effects of hypothermia developed, in an attempt to prolong the safe ischaemic time of the brain during cardiac arrest. This interest was further fuelled by numerous case reports involving patients submerged in icy water, buried in avalanches or trapped in crevasses. These patients achieved good neurological outcomes after slow rewarming despite prolonged arrest times [7, 8, 9]. Grewal and Thomas' case describes a rare example of this phenomenon in an urban setting [3].

During cardiac arrest, there is a sudden, massive drop in oxygen delivery. To avoid or minimise hypoxic tissue damage, increased delivery or reduced demand is required. Conventional cardiopulmonary resuscitation, comprising chest compressions and artificial ventilation, aims to maintain oxygen delivery. However, manual chest compressions produce, at best, only 20–30% of the usual cardiac output [10]. Mechanical compression devices, such as the Lund University Cardiopulmonary Assist System (LUCAS®, Stryker Medical, Portage, USA), represent a major recent advancement in cardiac arrest management. They provide consistent high‐quality chest compressions for prolonged periods, as demonstrated in Grewal and Thomas' case [3]. Additionally, they can be applied in transit, improving the safety of transfers to specialist intervention centres such as those providing primary percutaneous coronary intervention (PPCI) or extracorporeal membrane oxygenation (ECMO).

Unfortunately, evidence for the effectiveness of interventions to reduce oxygen demand has not been forthcoming. Despite the anecdotal success stories, active cooling did not significantly improve survival outcomes following cardiac arrest in two large international multi‐centre randomised controlled trials [11, 12]. The two targeted temperature management (TTM) trials found almost identical survival outcomes when comparing normothermia to hypothermia in a combined population of almost 3000 patients. They failed to even signal that active cooling confers benefit which contradicts the multiple cases of neurologically intact survival following both accidental and iatrogenic hypothermic cardiac arrest.

The reason for this contradiction perhaps relates to timing. In all the reported cases of accidental hypothermia, and when a deep hypothermic circulatory arrest is used during surgery, cooling precedes the reduction in cardiac output. Indeed, Grewal and Thomas describe this same sequence of events in their case [3]. It is possible that cellular protection is only effective if hypothermia is established before, or very rapidly after, circulation is arrested, otherwise irreversible hypoxic injury will ensue during even a moderate time of normothermia.

Inducing hypothermia within this time window presents a challenge because the human body contains a significant amount of thermal energy. Being predominantly composed of water which has a high specific heat capacity (4184 J.kg⁻¹.K⁻¹), a large amount of energy needs to be extracted to reduce body temperature. In the first TTM trial, the average time to reach the cooling target (33 °C) was over 8 h from collapse [11]. Subsequently, efforts were made to decrease the interval between circulatory arrest and achieving therapeutic hypothermia. In 2010, the PRINCE trial investigated pre‐hospital intranasal evaporative cooling to induce limited hypothermia specifically targeting the brain [13]. Average time to achieve the cooling target (34 °C) in the treatment and control groups was 102 and 291 min, respectively. The PRINCE trial demonstrated a non‐statistically significant increase in neurologically intact survival from 21.4% to 34.4% in a population of 180 witnessed cardiac arrests. A subgroup analysis, albeit underpowered, suggested an even greater survival increase from 17.6% to 43.5% when patients received resuscitation within 10 min of collapse.

The PRINCESS trial, a follow‐on study conducted in 2019, supported the findings described above [14]. In 677 patients presenting with witnessed out‐of‐hospital cardiac arrest (OOHCA), a good neurological outcome was demonstrated in 16.6% of the cooling group compared with 13.5% of the control group. Cooling was initiated approximately 20 min after the collapse. Median time to target temperature (<34 °C) was 105 and 182 min in the intervention and control groups, respectively. Both trials suggest that a good outcome following cardiac arrest relies on the effective delivery of several interventions in a timely fashion.

Grewal and Thomas note that the current best practice is to rewarm hypothermic patients extracorporeally using a cardiac bypass or ECMO circuit, if available [3]. This is advantageous in providing cardiac support in the event of further circulatory arrest or arrhythmias (common during the rewarming phase) whilst ensuring adequate oxygen delivery to minimise further neurological insult. However, these costly interventions require specialist expertise meaning that they are not provided in every hospital, including the location of the reported case. They are also not without risk. Complications can arise from cannula insertion (vessel damage, bleeding, failure to cannulate) or extracorporeal circulation (stroke, cerebral haemorrhage, disseminated intravascular coagulation, cardiovascular instability) [15]. Despite this, interest has increased significantly in using these interventions to manage refractory cardiac arrest and hypothermia. An ECMO circuit can meet the body's oxygen demands completely. It could therefore theoretically be used to support the transfer of patients in refractory arrest to specialist centres with definitive management such as PPCI for the restoration of native circulation. This has become more realistic in clinical practice with technological advancements producing smaller and more lightweight devices [16].

The international Extra‐Corporeal Life Support Organisation live registry database demonstrates a 30% survival rate in 14,839 cases of adult ECMO cardiopulmonary resuscitation [17]. As with other interventions during cardiac arrest, ECMO must also be implemented rapidly. Outcomes are significantly better if patients are put ‘on pump’ within 1 h of collapse [15, 16]. Currently, this is almost impossible for patients presenting with OOHCA, which has stimulated interest in the feasibility of providing ECMO in prehospital settings. In 2017, a group in Paris evaluated this and improved their average ‘on‐pump’ time from 90 to 70 min, while neurologically intact survival increased from 8% to 29% [18]. Several other trials are now investigating similar objectives. The London‐based Sub30 trial aims to be ‘on pump’ within 30 min [19], whilst the Dutch ON‐SCENE trial is investigating the implementation of prehospital ECMO nationally [20].

Interpreting and performing cardiac arrest research remains challenging. Study interventions are confounded by the timing and events surrounding their application. There is the potential for interventions with real benefit to perform poorly in trials because of study design and external factors. However, the benefit of reducing time to critical intervention is universally agreed. For patients presenting with OOHCA, this will almost certainly mean the commencement of treatments before hospital arrival. To achieve real progress in this area, collaboration between researchers, hospitals, ambulance services and local helicopter emergency medical services is vital.